Drones and Data in the Battle Against Wildfires

Few challenges test the limits of technology as severely as natural disasters. The day when humanity is fully prepared for them never seems to arrive. Even the most obvious assumptions can prove unreliable. Albrecht Beck, founder and CEO of Prepared International, UN adviser on natural disaster recovery and a regular speaker at the Disasters Expo, recalls that many lives were spared during Hawaii’s devastating 2023 wildfires precisely because the warning systems were not activated.

“In the case of the fires in Hawaii, there was no warning because the only disaster people there recognise as the primary threat has always been the tsunami,” Beck explains. “So, had an alert been issued, people would have fled towards the mountains—straight into the path of the fire.”

Collaboration is essential when facing a natural force capable of wreaking record levels of destruction. In California alone, damages this year have been estimated at an unprecedented $165 billion. According to the European Commission’s Joint Research Centre, over 60,000 fires ignite across the EU each year, consuming an average of 500,000 hectares, claiming lives and generating losses of around €2 billion.

On 17 June, G7 leaders signed the Kananaskis Forest Fire Charter, marking a landmark commitment. The deployment of fire management technologies features prominently in this declaration. In parallel, initiatives such as the European Patent Office’s (EPO) Firefighting Technologies platform have been launched to gather the most promising ideas and make them available to companies and public authorities.

Prevention is key

It is striking that a reconstruction specialist such as Albrecht Beck places so much emphasis on the pre-disaster phase: “Afterwards, it is always too late. We must succeed during preparation, not once disaster has already struck on such a dramatic scale,” he insists.

When it comes to forest fires, prevention remains the major unresolved challenge in every country. Public spending on firefighting still outweighs prevention budgets by as much as six to one. The OECD has warned of the common practice of “fire loans”, in which funds earmarked for prevention are diverted to finance emergency response and recovery.

Innovation in wildfire management must be pursued from multiple disciplines. The Forest Science and Technology Centre of Catalonia, Spain, has launched 11 Living Labs as part of the FIRE-RES project. These open-air test sites allow firefighters, researchers, farmers and local communities to trial new approaches, including drone-based early detection systems, controlled burns and fire-resistant construction materials.

Meanwhile, Canada’s FPInnovations’ Wildfire Operations Research programme has employed industrial-grade cinema smoke machines to train fire-alert technologies in recognising wildfire signatures. The datasets generated are now available to technology developers keen to refine their artificial intelligence (AI) systems.

Data as a fireline

A digital ecosystem of data platforms will be indispensable in understanding and combating wildfires in the years ahead. The EU has launched the EFFIS programme, while the United States’ ACERO project is harnessing NASA’s expertise to develop an air traffic management system enabling drones to fight fires even in low light and poor visibility. ACERO has also acquired a portable airspace management system (PAMS), compact enough to fit in a briefcase, to support drone operators in the field.

Reliable real-time data is crucial. EURO1k is the first numerical weather model covering Europe and parts of North Africa with a one-kilometre resolution. Its developers claim it can accurately simulate small-scale weather events, including thunderstorms, hail and gales.

Satellite systems can now detect a fire in as little as one minute, while the ALERT network, run by the universities of Reno, Nevada and Oregon, has confirmed detections in under three minutes using cameras and sensors. Spain’s Technosylva combines geospatial and government data with GPS feeds from firefighters to generate real-time simulations guiding operational decisions. Its software, used by power companies, manages over 20,000 incidents worldwide each year.

Bringing detection and suppression technologies together is proving decisive. FireMap, developed by WIFIRE Lab—a spin-off of the San Diego Supercomputer Centre—uses AI to produce predictive maps of wildfire trajectories within minutes. Meanwhile, Stanford University researchers have developed a fire-retardant gel that acts as a protective barrier for forested areas..

Unmanned aviation takes on wildfires

Alongside the growth of data platforms, one of the most dynamic areas of innovation lies in unmanned aviation. China has begun mass production of the AG600 Kunlong, the world’s largest amphibious aircraft, developed by the Aviation Industry Corporation of China (AVIC). With a range of 4,500 kilometres, it can carry 12 tonnes of water for firefighting missions.

Drones may well prove the most effective tool for early fire mitigation. The Windracers Ultra fixed-wing drone can deliver over 100 kilograms of fire retardant and is fitted with artificial intelligence designed by the University of Sheffield. Using thermal and optical imaging, it can autonomously detect and evaluate wildfires.

The Ignis system, a funnel-shaped attachment fixed to the underside of a drone, can release 450 small incendiary capsules, known as “dragon eggs”, in about four minutes. Each contains two chemicals that ignite upon impact, creating controlled burns to deprive a larger fire of fuel.

Elsewhere, a Canadian company is developing quadcopter drones for wildfire suppression. Equipped with sensors and AI-driven swarm algorithms originally designed for defence, each drone is capable of lifting 400 kilograms—about a third of a conventional helicopter’s load—and can operate at night.

“Today’s technology can handle vastly greater volumes of data,” concludes Albrecht Beck. “Drone swarms can now provide continuous, real-time coverage, and artificial intelligence makes sense of it all. Just four or five years ago, it was overwhelming—we could not process it quickly enough during an emergency. Now, AI does that work for us.”

Sources:

• FPInnovations
• UNDRR
• Joint Research Center
• OCDE